CA2085180C - Nitrogen adsorption with a ca and/or sr exchanged lithium x-zeolite - Google Patents

Abstract

The present invention is directed to a process for separating nitrogen from gas mixtures containing nitrogen and less strongly adsorbed components such as oxygen, hydrogen, argon or helium by use of an at least binary exchanged X-zeolite having lithium and calcium and/or strontium ions in ratio of preferably 5% to 50% calcium and/or strontium and 50% to 95% lithium.

Description

~ ~1$ ~ 8 NITROGEN ADSORPTION WITH A CA AND/OR SR EXCHANGED
LITHIUM X-ZEOLITE

TECHNICAL FIELD
The present invention is directed to gas separations usin~
nitrogen selective adsorbents. More particularly, the present invention is directed to at least binary exchan0ed X~zeolltes using a combinatlon o~ lithium and calclum and/or strontlum cations to recover oxy0en or nitro~en from ~as mixtures containing them, such as alr.
BACKGROUND OF THE PRIOR ART
Adsorptlve separations usln~ zeolitic structures as adsorbents are well known in the prior art for resolvin~ a multitude of 0as mixtures. Such separations are predicated upon the compositions of the gas mixtures and the componants' selectivity for adsorption on adsorbents, such as zeolites.
rhe use of nitro0en in industrial ~as applica~ions has seen significant growth particularly with the development of non-cryogenic gas mixture separations. A major field of nitrogen separation comprises the separation of nitro~en from air. The removal of nitro~en from air results in an enriched oxy~en gas component which is less strongly adsorbed by appropriate zeolites which are selective for nitrogen adsorption. When oxygen is desired as product typically a~ elevated pressure, it is desirable to adsorb nitro~en from atr to result ln unadsorbed oxy~en enriched product passing over a nitrogen selectlve adsorbent. The nitro~en is then removed during a sta~e of desorption, typically at lower pressure. This results in oxy~en being recovered at the pressure of the feed air, while nitro~en ls recovered at a pressure below the feed air pressure. As a result, for the production of oxygen without si~n1ficant pressure loss in an adsorptive separation of air, it is desirable to utilize nitro~en select.tve adsorbents such as the family of 7eolites.
Although various zeolites are naturally occurring and various synthetic zeolites are known, some of which have appropriate selectivities for nitro~en over oxygen and other less ,:~

! 2 s~rongly adsorbed substances such as hydrogen, ar~on, helium and neon, the industry has attempted to enhance the performance oF
various zeolites to improve their selectivi~y and capacity for nitrogen over such less strongly adsorbed substances such as oxygen. For ins~ance, in U.S. Patent 4,481,01~, various sodium or calcium X-zeolites and faujasites are known which have low sllicon to aluminum ratios in the order of approximately 1 to 1.2. The zeolites of this patent have utility for nitrogen adsorption, particularly from gas mixtures such as air when activated in a par~icular technique which minimizes the presence of water as it evolves from the material. The technique ls further described in U.S. Patent 4,544,378.
In U.K. Patent 1,580,928, a process for makin~ low silica X-zeolites ("LSX"; where LSX is X zeolite with a Si/Al = 1) is set forth comprising preparing an aqueous mixture of sources of sodium, potassium, aluminate and silicate and crystallizing th~
mixture at below 50~C or aging -the mixture at 50~C or below followed by crystallizing the same at a temperature in the ran~e of 60~C to 100~C.
Gunter H. Kuhl in an article "Crystallization of Low-Silica Faujasite" appearing in Zeolites (1937) 7, p451 disclosed a process for making low silica X-zeolites comprising dl~solvin~
sodium aluminate in water with the addition of NaOH and KOH.
Sodium silicate was diluted with the remainin~ water and rapidly added to the NaAlO2-NaOH-KOH solwtion. Th0 ~elled mixture was then aged in a sealed plastic jar for a specified time at a specified temperature. The product was filtered and ~ashed.
Other low silica X-zeolite synthesis processes are available, such as those set forth in U.S. Patent 4,606,899.
In U.S. Patent 3,140,931, the use of crystalline zeolitic molecular sieve material having apparent pore sizes of at least 4.6 Angstroms for separating oxygen-nitrogen mlxtures at subambient temperatures is disclosed.
U.S. Patent 3,140,932 specifically claims Sr, Ba, or Nl ion exchanged forms of zeolite X.
U.S. Patent 3,313,091 claims the use of Sr X-zeolite at adsorption temperatures near atmospheric, and subatmospheric desorption pressures.
It is also known in U.S. Patent 4,557,736 to modify X-zeolites by ion exchange of available ion sites with several divalent cations to produce a binary ion exchanged X-zeolite wherein the binary ions which are exchan~ed comprise calcium and strontium. These binary ion exchan~ed X-zeolites using calcium and strontium are reported to have hi~her nitrogen adsorption capacity, low heat of nitrogen adsorption and good nitrogen selectivlty for air separation.
It is also known to exchange X-zeolites with lithium to provide an improved nitrogen selective adsorbent as set forth in U.S. Patent 4,859,217. This patent sulggests an improved nitrogen adsorbent can be achieved when an X-zeolite is exchan~ed with lithium cations at greater than 88%. The startin~ material for this patented zeolite is sodlum X-zeolite. Therefore, the patent recites a lithium-sodium X-zeoli~e for nitrogen adsorption.
The prior art lithium X-zeolite was reported in U.S. Patent 3,140,933 as useful for ni~ro0en-oxygen separations.
In an article entitled, "Investigations of the Arrangement and Mobility of Li ions in X- and Y-zeolites and the Influence of Mono- and Divalent Ca~ions on It" by H. Herden, W. ~. Einicke, R.
Schollner and A. Dyer, appearing in J. Inorganic Nuclear Chemistry, Vol. 43J No. 10, pages 2533 thru 2536 (1981), the existence of mixed cation, lithium and calcium exchan~ed X-zeolites are set forth. Physical parameters of the exchan~e zeolites are discussed with a general recitatlon to adsorptive and catalytic utilities of zeolites in general.
Although improved exchanged X-zeolite adsorbents have been reported in the art for nitrogen adsorptions, and particularly the hi~h performance of hi~hly lithium exchange X-zeolites are known, such zeolites are difficult to achieve at hi~h level lithium exchange and consti~ute an expenslve adsorbent to produce for nitrogen separations. Such production difficulties and expense limit the use of such exchanged X-zeolites to produce either nitrogen or oxygen in competltion with other separation technologies, such as cryogenic distillation and membrane separations. Therefore, a problem exists in the ar-t for providing an appropriately exchanged X-zeolite for effective nitrogen adsorptive separation using an exchan~ed X-zeolite which is readily produced ar,d has a favorable cost so as to result in competitively priced nitrogen, oxygen or other gas component product pricing. The art also desires to have a hi~h selectivity exchanged X-zeolite with reasonable working capacities which do - 4 - s~,~ 8 ~ ~3 not inhibit continuous operation or adsorbent re~eneration.
These unresolved problems are achieved by the present invention, which is set forth below.

BRIEF SUMMARY OF THE INVENTION
The presert invention is a process for selectively adsorbin0 nitrogen from a gas mixture containing nitro0en and at least one less strongly adsorbed component which comprises contacting the gas mixture with an adsorbent which is selective for the adsorption of nitrogen, comprisiny a crystalline X-zeollte having a zeolitic Si/Al ratio less than or equal to 1.5 and an at least binary ion exchange of exchangeable ion content wlth between 5%
and 95% lithium and with between 5% and 95% of a second ion selected from the group consisting of calcium, strontium and mixtures thereof, wherein the sum of the llthium and second ion in ion exchange is at least 60% of the exchan~eable ion content.
Preferably, the zeolite is ion exchanged with llthlum to approximately 50% to 95%.
Preferably, the zeolite is ion exchanged wlth the second ion to approximately 5~ to 50%.
Preferably, the zeolite is ion exchanged with approximately 15% of the second ion and 85% lithium.
Preferably, the second ion is calcium. Alternatively, the second ion is strontium.
Preferably, the zeolite is ion exchan0ed with approximately 15% calcium and 85% lithium.
Preferably, the gas mixture contains nitro~en and oxygen.
More preferably, the ~as mixture is air.
Preferably, the Si/Al ratio is approximately 1.
Prefera~ly, an oxyg0n and nitrogen containing gas mixture contacts a zone of such adsorbent, the nitrogen is selectively adsorbed and the oxy~en passes through the zone and is recovered as an oxygen enriched product.
Preferably, the oxygen product has a purity of at least approximately 90% oxygen.
Preferably, the adsorption is conducted at an average bed temperature in the range of approximately 55 to 135~F.
Preferably, the zone is operated through a series of steps comprising: adsorp-tion, during which the gas mixture contacts the adsorbent, nitrogen is selectively adsorbed and oxygen passe-s .

through zone as product; depressurization during whlch the gas mixture contact is discontinued and the zone is reduced in pressure to desorb the nitroyen; and repressurization with oxy~en product to the adsorption pressure.
Preferably, the adsorption pressure is in the range of approximately 35 to 65 psia.
Preferably, the depressurization is conducted down to a level in the ran~e of approximately 14.7 to 16.7 psia.
Alternatlvely, the zone is operated through a series of steps comprising: adsorption, during wh~ch the gas mixture contacts the adsorbent, nitrogen is selectively adsorbed and oxygen passes through zone as product; depressurization during which the gas mlxture contact is discontinued and the zone is reduced ln pressure to desorb the nitroyen; evacuation to further desorb the nitrogen to below ambient pressure; and repressurization with oxygen product to the adsorption pressure.
Preferably, ~he adsorption pressure ls in the range of approximately 900 to 1600 torr.
Preferably, the evacuation is conducted down to a level in the range of approximately 80 to 400 torr.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a graph of nitrogen capaclty for the extent of lithium exchan0e in a lithium, sodlum LSX-zeolite at 1 a~mosphere 2~ and 23~C showing that capacity uniformly is less than what mi~ht be expected from straight line extrapolation of 100% sodium to 100% lithium ion content.
FIG. 2 is a ~raph of nitrogen capacity for the extent of calcium exchange in a calcium, sodium LSX-zeolite at 1 atmosphere and 23~C showing that capacity uni~ormly is less than what mi~ht be expected from strai~ht llne extrapolation of 100% sodium to 100% calcium ion content.
FIG. 3 is a ~raph of nitro~en capacity for the ex~ent of lithium exchangç in a lithium, calcium LSX-zeolite at 1 atmosphere and 23~C showing that capacity uniformly and unexpectedly is in excess of what might be expected from strai~ht line extrapolation of 100% calcium to 100% lithium ion content.
FIG. 4 ls a ~raph of nitrogen capacity for the extent of lithium exchange in a lithium, strontlum LSX-zeolite at 1 atmosphere and 23~C showing that capacity uniformly and - 6 - 2l~8 ~

unexpectedly is in excess of what might be expected from stralght line extrapolation of 100% strontium to 100% lithium ion content.
FIG. 5 is a graph oF isothermal nltro0en workin~ capacity for the extent of lithium exchange in a lithlumJ calclum LSX-zeolite and lithium, sodium LSX-zeolite from 0.2 to 1.0 atmosphere and 23CC showin~ tha~ workin0 capacity for lithium, calcium LSX-zeolite uniformly and unexpectedly is in excess of what might be expected from straight line extrapolation of 100%
calcium to 100% lithium ion content, in contras~ -to lithlum, sodium LSX-zeolite, which is well below its expected workin~
capacity.
FIG. 6 is a graph of isothermal nitrogen working capacity for the extent of calcium exchange in a lithium, calcium LSX-zeolite and calcium, sodium LSX-zeolite from 0.2 to 1.0 atmosphere and 23~C showing that working capacity for lithiwm, calcium LSX-zeolite uniformly is in excess of what might be expected from straight line extrapolation of 100% calcium to 100%
lithium ~on content, in contrast to calcium, sodium LSX-zeolite, which is below or even to its expected working capacity.
FIG. 7 ls a graph of binary N2/02 selectivity calculated by IAST for air feed at 1.45 atmospheres at 30~C for the extent of lithium exchange in a lithium, calcium LSX-zeolite and lithium, sodium LSX-zeolite showing that the binary N2/02 selectivities for lithium, calcium LSX-zeolite are higher than the selectlvities for lithium, sodium LSX-zeolite at the same lithium exchange level.
FIG. 8 is a graph of binary N2/02 selectivity calculated by IAST for air feed at 1.45 atmospheres, 30~C for the extent of calcium exchange in a lithium, calcium LSX-zeoli~e and calcium, sodium LSX-zeolite showing tha~ the binary N2/02 selectivitles for lithium, calcium LSX-zeolite are higher than the selectivities for sodium, calcium LSX-zeolite at the same calcium exchange level.
FIG 9 is a graph of the effect of llthium exchange leve1s on nitrogen loading a~ 700 ~orr and 23~C for binderless X-zeolite with a silicon to aluminum ratio of 1.2 showing that the X-zeolites are similar to the low silicon X-zeolites (LSX) and also display the unique and unexpected performance of the present invention .

.. . ~ . : : :~
..
, , .

- 7 - % ~3 8 ~

DETAILED DESCRIPTION OF THE INVENTION
The process of the present invention directed to nltroaen adsorption from gas mixtures of less strongly adsorbed components, such as: oxyyen, hydrogen, argon and helium is achieved by the use of a binary, ternary or further exchanged X-zeolite wherein, ~ypically, a sodium or sodium, potassium X-zeolite is exchanged with lithium and calcium and/or strontlum elther co-currently or sequentially to result in a lithiuml calcium and/or strontium X-zeolite, whlch may contain a residual 19 minor amount of sodium or potassium ions. The lithlum content is in the range of approximately 5% to 95% lithium, preferably 50%
to 95~, more preferably, 85%. The approprlate respective calclum and/or strontlum content is between 5% and 95% calcium and/or strontium, preferably 5% to 50%, more preferably 15% calclum, but obviously the combination of lithium and calcium and/or strontium chosen for any set of percentages would not excPed 100% and in some instances may be less than 100% based upon residual sodium or potassium cation content. Preferably, the X-zeolite is a low silica X-zeolite with a Si/Al ratio o~ approximately 1 and with approximately 15% calcium and 85% lithium, although any combination of exchange wherein the lithium and calcium and/or strontium is at least 60% of the exchangeable ion content in the zeolite is acceptable.
Although other ion ~orms of X-zeolites can be used, typically a sodium or mixed sodium/potassium X-zeolite is used to prepare the at least binary ion exchanged materials. Typically, ~he ions are exchanged co-currently, al-though they can be exchanged sequentially, for example by exchanging a sodium X-zeolite with calcium to produce a calcium X-zeolite, which can then be partially ion exchanged with lithium to yield the desired adsorbent. The ion exchange is accomplished by contacting the sodium or mixed sodium and potassium X-zeolite with a salt of the ion to be exchan~ed. Other methods o-f ion exchange are contemplated and can be used for the present invention.
These binary ion exchanged X-zeolite adsorbents have higher nitrogen working capacity than pure calcium exchanged X-zeolites and comparable nitrogen working capacity to lithium exchanged X-zeolites, yet have heats of adsorption for nitrogen which allow for commercial utility. Additionally, the at least binary ion exchanged X-zeolites of the present invention exhibit higher - 8 .

nitrogen/oxygen selectivity than that observed for the prlor art lithium, sodium X-zeolite at the same lithium exchange level and higher than that observed for the prior art calcium, sodium X-zeolite at the same calcium exchange l~evel.
An added benefit of the lithium, calcium X-zeolites of the present invention is that they do not display the deterioration in performance with increasing temperature observed for the prior art lithium, sodium X-zeolites ln vacuum sYJin~ adsorptlon processes.
The use of calcium to make lithium, calclum X-zeoli~es results in a lower cost adsorbent than the hi~hly exchan0ed levels of lithium X-zeolite because calcium salts cost less than lithium salts, and the exchange of calcium for sodium is much more thermodynamically favorable than the exchan~0 of lithium for sodium. The ability to alter the respective amounts of calcium and/or strontium and lithium exchange provides far more flexibility in optimizing the adsorbent properties for various gas separation operations. A preferred use for the at least binary ion exchanged X-zeoli;tes of the present invention is the separation of nitrogen from oxygen in air using a pressure swing adsorption ("PSA") or vacuum swing adsorption ("VSA") process.
In such a process, an adsorbent bed comprisiny binary ion exchanged lithium, calcium X-zeolite, as described above, is initially pressurized with oxygen. A ~as stream comprising nitrogen and oxygen, such as air at a temperature between 0~ and 50~C and a pressure between 1 atmosphere and 5 atmospheres, is passed over the adsorbent bed. A portion o~ the nitro~en in the gas stream is adsorbed by said ion exchan~ed zeolites, thereby producing an oxygen-enriched product stream. The nitrogen containing adsorbent bed is subsequently depressurized and evacuated with the option of being purged with oxy~en enriched gas to produce a nitrogen enriched stream. The bed is then repressurized with product oxygen and adsorptlon can be reinitiated. Alternatively, these materials can be used for recovering a nitrogen enriched product using, for example, an existing nitrogen vacuum swing adsorption process as described in U.S. Patent 4,013,4~9, wherein the proc~ss includes the steps of feed, rinse, desorption, and repressurization.
Although the at least binary exchan~e levels of lithium and calcium and/or strontium on the X-zeolite demonstrate high .

.

- 9~ 8~

performance for nitrogen selective adsorptive separation, additional beneFit can be achieved by the appropriate selection or treatment of the aluminum con~ent of the zeolitic framework to produce preferred results. X-zeolites typically have a silicon to aluminum ratio less than or equal to 1.5 and typically betwe0n 1.2 and 1.5. 'For the purposes of the present invention using binary excha,n~ed X-zeolites however, it is preferred to use a low silica X-zeolite havin~ a silicon to aluminum ratio oF
approximately 1.
The a~sorbent must be dehydrated before bein~ used For ~as separation usin~ a thermal activation step. Such a thermal activatlon step can be achieved by a number of different methods in whic,h the zeolitic water and the hydration spheres are carefully removed and the amount of water in the ~aseous environment in contact with the zeollte during this step is minimized. That ls, the partial pressure of water making such contact should be less than about 0.4 atmospheres, preferably not more than about 0.1 atmospheres.
One method of accomplishing this is to sub~ect the at least binary exchanged X-zeolite composltion, which contains up to about 30% by wei~ht of water, to pressures in the ran~e of about 0.1 to lO atmospheres while maintaining sufficient molar mass velocities and residence times of a flow of a non-reactlve pur0e yas, that is a molar mass velocity of about 0.5 to 100 kiloarams mole per meter squared hour and a residence time of no greater than about 2.5 minutes, and then heat the composition at a ~emperature ramp of 0.1 to 40~C per minute up to a temperature of at least about 300~C and no greater than about 650~C. The residence time is defined as the volume of the column o,r other unit used to thermally activate the zeolite divided by the volumetric flow rate of the purge gas at the standard temperature and pressure. The molar mass velocity is the flow rate of the purged gas divlded by the cross-sectional area of the column used for thermal activation. The purpose of the purge gas is to provide a sufficient mass for efficient heat and mass transfer from the surface of the adsorbent at a residence time to limit the water in the purge gas exiting the adsorbent bed to the desired low limits. The minimum residence time is determined by economic and process constraints, althou~h times of less than 0.0025 minutes would appear to provide no advantages.

.
; ~ :
.
, .
. ~. . .. .

- 10 - ~ 8~

Ano-ther method o-F thermal actlvatlon is to conduct the activation under less than about 0.1 a-tmospheres vacuum without the use of the purge gas and to heat the material to the desired activation temperature and a ramp temperature of from 0.1 to 40~C
per minute.
Still another method that is available for thermal activation of zeolitic adsorbents is the use of microwave radiation, conditions that are described in U.S. Patent 4,322,394, of which the description of the microwave procedure for thermally activating zeolites is incorporated herein by reference.
We have found unexpectedly that the nitrogen capacities of the at least binary lithium, calclum and/or strontium ion exchanged forms of X-zeolite unexpectedly exceed what might be expected from a straight line extrapolation of the capacities of the two end member ions. In contras-t, the nitrogen capacitles of the prior art binary calcium, sodium ion exchange forms of X-zeolite and the binary lithium, sodium ion exchan~ed forms of X-zeolite are always less than what mi~ht be expected from a straight line extrapolation of the nitrogen capacities of the two end members. The unexpected performance of mixtures of lithium and calcium is also observed for ternary systems containin~, for example, residual sodium or potassium, as long as the sum of the lithium and calcium and/or strontium exchange levels is greater than about 60%.
In the ion exchange experiments set forth below which demonstrate the present invention, various zeolite startin~
materials were used. Sodlum X-zeolite powder with a Si/Al ratlo of 1.2 was obtained from the Llnde Division of Union Carblde Corporation.
Sodium, potassium LSX-zeolite powder was prepared by the method of Kuhl and Sherry in UK 1,580,928. In ~hat patent, a process for making low silica zeolites is set forth comprisin~
preparin~ an aqueous mixture of sources of sodium, potassium, aluminate and silicate and crystalllzing the mixture at below 50~C or a~ing the mixture at 50~C or below followed by crystallizing the same at a temperature in the ran~e of 60~C to 100~C. (See also Kuhl,G.H.7eolites 1987, 7, 451) Other samples of sodium, potassium LSX-zeolite powder were prepared from clay , :

11 - 2 ~

by the process of Butter et al. in US 4,606,899 in which kaolin clay, calcined to at least 700~Cj is converted to LSX-zeolite by agitating a reaction mixture, comprised o~ the clay with sodium and potassium hydroxide, at temperatures in excess of 50~C and seeding the resulting mix~ure wlth LSX-zeolite at a predetermined time after the reaction has been initlated.
, .

Lithium, Sodlum LSX-zeollte Control Lithium LSX-zeolite was prepared by ion exchange of sodium, potassi4m LSX zeolite powder using five statlc exchan~es a~t 100~C
with a 6.3-fold equivalent excess of 2.Z M LiCl. Various exchange levels of lithium, sodium LSX-zeolite wers prepared by addin~ nine separate samples of the inltially prepared lithium LSX-zeolite powder ~o various amounts of 0.1 M NaCl and stirring at room temperature for about 4 h. The mixed catlon samples were filtered but not washed to prevent hydrolysis of the llthium cations. The use of a dilute solution made the errors in cation levels introduced by the solution retained on the fllter cake insignificant.
The samples were analyzed by Inductively Coupled Plasma-Atomic Emission Spectroscopy (ICP-A~S) for silicon and aluminum and Atomic Absorption Spectroscopy (M ) for lithium, sodium, and potassium. Table I contains the results of elemental analyses for the amount of lithium and sodium in the exchanged samples.
Adsorptive capacities for nitrogen (N2) were obtained using a conventional McBain gravimetric adsorptlon unit that could contain nine samples. Samples were first superficially dried at 110~C in an oven purged with N2 at a hi~h flow rate.
Approximately 5 g were loaded into the McBain sample buckets, and the samples were heated under vacuum at 1~C/min or less to 550~C.
The samples were held a~ ~50~C until the pressure dropped to about 10 microns of Hg. After activation, N2 isotherms were obtained to 1 atm at 23~C. The isotherm data was fit to the standard Langmuir :isotherm equa~ion:

Nm = mbP/(1~bP) - 12 - %~8~

where Nm is the amount adsorbed, P is the pressure, m is the monolayer capacity and b is the affinlty parameter. The fits were used to generate N2 capacities and isothermal N2 worklng capacities reported in Table I.
FIG 1 compares the observed N2 capacities for the extent of lithium exchange level in lithium, sodium LSX-zeolite to what might be expected from straight line extrapolation of 100% sodium to 100% lithium ion content. It shows that for lithium, sodlum LSX-zeollte, N2 capacity.uniformly is ]Less than wha-t might be expected.
The effect of lithium exchange level on W2 capacity for lithium, sodium binary exchanged forms of LSX-zeolite ls very similar to that reported for X-zeolite by Chao in US 4,859,217.

::

- l3 %~

TABLE I
Nitrogen Capacities for Mixed Cation (Li,Na)LSX a~ter Activation ~o 550~C, 2xlo 2 torr sample Li/Al Na/Al N~obs)1, N~delta),2 numbereq ratio eq ratio mmol/g mmol/g 1 1.03 o.ol 1.35 o.so 2 0.90 o.1o 1.06 0.70 3 0.83 0~20 0.74 0.51 4 0.70 0027 0.47 0.32 0.6~ 0.34 o.~o 0.28 6 0.58 0.45 0.~2 0.29 7 0.43 0.55 0.42 0.29 8 0.30 0.66 0.39 0.26 9 0.21 0.75 0.39 0.26 lo o.ll 0.86 0.44 0.31 11 n/a 1.00 0.43 0.30 1Nm(obs) = nitrogen capacity at 1 atm. and 23~C.
2Nm(delta) = isothermal working capacity from o.2 to ~.o atm at 230C.
n/a = not analyzed Calcium, Sodium LSX-Zeolite Control Sodium LSX-zeolite was prepared by ion exchange of sodium, po~assium LSX-zeolite using three static exchan0es at 100~C with a 4.2-fold equivalent excess of 1.1 M NaCl. Various exchange levels of calcium, sodium LSX-zeolite were prepared by adding nine separate samples of the initially prepared sodium LSX-zeolite powder to varying amounts of 0.05 M CaCl2 and stirring at room temperature for about 4 h. The mixed cation samples were filtered but not washed. Table II contains the results of elemental analyses for the amount of calcium and sod~um in the exchanged samples. N2 capacities and isothermal -. ~

1 ' ' ' .~ . .

. ~ ". ;;

working capacities were obtained at 23~0 usin~ the McBain ~ravimetric adsorption unit as described in Example 1.
FIG 2 compares the observed N2 capacities for the extent of calcium exchange level in calcium, sodlum LSX-zeolite to what mi~ht be expected from stral~ht line extrapolatlon of 100% sodium to 100% calcium ion conten~. It shows that for calcium, sodlum LSX-zeollte, N2 capacity uniformly i5 less than what mi~ht be expected.
The effect of calcium exchan~e level on N2 capac~ty for calcium, sodium binary exchan~ed forms of LSX-zeolite is very similar to that reported for the effect of calcium exchan~e level on N2/02 selectivity for X-zeolite by Coe et al. ln US 4~481,018.

- 15 - s~

TABLE II
Nitrogen Capacities for Mixed Cation (Ca,Na)LSX after Activation to 550~C, lx10-2 torr _ _ _ _ _ _ _ _ sample Na/Al Ca/Al N~tobs)2~ N~(delta),3 number eq ratio eq ratio mmo:L/g mmol/g 1 l.oo n/a 0.4:3 0.31 2 0.86 0.10 0.5:3 0.36 3 0.77 0.19 0.49 0.33 4 0.70 0.29 0.46 0.35 0.58 0.38 0.53 0.38 6 0.50 0.46 0.65 0.46 7 0.36 0.56 0.85 0.56 8 0.30 0.66 1.02 0.63 9 0.25 ~.72 1.14 0.67 0.20 0.77 1.14 0.65 111 n/a 0.97 1.53 0.74 lVacuum activation to 400~C, ~lx10-5 torr 2Nm(obs) = nitrogen capacity at 1 atm. and 23~C.
3Nm(delta) = isothermal working capacity from 0.2 to 1.0 atm at 23~C.
n/a = not analyzed Lithlum, Calcium LSX Zeolite N~ Capacities Various exchange levels of lithium, calcium LSX-zeolite were prepared by adding six separate samples of lithium LSX-zeolite powder to varying amounts of 0.05 M CaCl2 and stlrring at room temperature for about 4 h. The mixed cation samples were filtered but not washed. Table III contains the results of elemental analyses for the amount of lithium and calcium in the exchanged samples and N2 capacities and isothermal working capacities at 23~C obtained using ~he McBain gravimetric adsorption unit as described ln Example 1. Binary lithlum, calcium LSX-zeolite samples with high levels of calcium lost - 16 ~ o ~

crystallinity as a result of drylng and activatlon; consequently, results from the high pressure volumetric unit ~described below and summarized in Table VI) obtained on the samples with hi0h calcium exchange levels were used for comparison to the controls.

FIG 3 compares the observed N2 capaclties for the extent of lithium exchange level in lithium, calcium LSX-zeollte to what might be expected from straight line extrapolation of 100%
calcium to 100% llthlum ion content. In contrast to the prior art binary lithium, sodium and calclunl, sodium exchanged zeolites, as demonstrated in Example 1 and Example 2, the observed N2 capacities for binary ion exchan~ed lithlum, calclum LSX-zeolite are uniformly and unexpectedly in excess of what might be expected. Comparison of FIGs 1, 2 and 3 demonstrates the improved performance of lithium, calcium blnary ion exchan~ed forms of LSX-zeolite over other binary ion exchanged forms of LSX-zeolite containing calcium or lithium known in the prior art.

' - 17 - s~8 TABLE III
Nitrogen Capacities for Mixed Cation (Li,Ca)LSX after Acti~ation to 550~C, lxlO 2 torr sample Li/Al Ca/Al N~(obs)2, N~(delta),3 number eq ratio eq ratio m~lol/g mmol/g 1 1.03 n/a 1.35 0.90 2 0.90 0.10 1.42 0.89 3 0.83 0.20 1.~9 0.93 4 0.73 0.30 1.45 0.89 5 0.59 0.41 1.48 0.87 610.48 0.51 1.53 0.90 710.28 0.72 1.59 0.~5 81 n/a 0.97 1.53 0.74 1Vacuum activation to 400~C, <lx10-5 torr 2Nm(obs~ = nitrogen capacity at 1 atm. and 23~C.
3N~(delta) = isothermal working capacity ~rom 0.2 to 1.0 atm at 23~C.
n/a = not analyzed N2 capacities at 23~C and 1 atm. of about 1.5 mmol/g for lithium, calcium ISX-zeolites with compositions around 70% lithium/30% calcium (FIG 3) are particularly unexpected. The prior art lithium, sodium LSX-zeolites would suygest that any LSX-zeolite cont~;nin~ 70% lit~ium should have a capacity of only 0.4 mmol/g (see 70% lithium in FIG1). hikewise, the prior art ~alcium, sodi~m LSX-~eolites would suggest that an LSX-zeolite containing 30%
calcium should have a capacity of only about 0.45 mmol/g (see 30% calcium in FIG 2).
The N2 capacity alone is not a measure o~ an adsorbent's ability to effect a separation of N2 from other components. Berlin, in US 3,3l3!09l, points out the - ~ ,. .................. ..

,. . :

.

~: :

- 18 - ~8~0 importance of the shape and slope of the component isotherms in the pressure region of interest. Consequ~ntly, the isothermal N2 working capacities from o.2 to l.o atm, a pressure region of interest for ~2 VSA air separation processes, were also compared for binary lithium, calcium LSX-zeolite from Example 3 and the comparable lithium, sodium LSX-zeolite and calcium, sodium LSX-zeolite controls in Examples 1 and 2. The isothermal N2 working capacity ~or lithium, calcium LSX-zeolite at any lithium exchange level is always higher than the working capacity for the prior axt lithium, sodium LSX-zeolite at the same lithium e~ch~nge level (Table III Nm delta contrasted with Table I N~ delta) as depicted in FIG 5. Likewise, the working capacity for lithium, calcium LSX-zeolite at any calcium exchangP level (Table III) is always higher than the working capacity for the prior art calcium, sodium LSX-zeolite at the same calcium exchange level (Table II) as depicted in FIG 6.
In addition, it can be observed that the nitrogen working capacities remain relatively constant for lithium exchange levels greater than about 50%. This suggests that the preferred composikion range for lithium, calcium LSX
zeolite for ~2 VSA is from 50-95% lithium and 5-50~ calcium.

2~ ~X~MP~ 4 Lithium Calcium Sodium LSX-Zeolite N2 Ca~acities The prior examples illustrated the improved performance of binary ion exchanged lithium, calcium LSX-zeolite over other binary ion exchanged forms of LSX-z~olite containing lithium or calcium. This example demonstrates the beneficial effect of lithium in combination with calcium for ternary ion exchanged forms that contain other cations, such as sodium, in addition to lithium and calcium.
Various exchange levels of lithium, calcium, sodium LSX-zeolite containing about 85% lithium were prepared by adding a number of s~parate samples of lithium LSX-zeolite powder to varying amounts of 0.1 N solution containing various amounts of CaC12 and NaCl and stirring at room temperature for about 4 h. The samples were filtered but not washed. Table IV-A contains the results of elemental analyses for the amounts of lithium, calcium and sodium in the exchanged samples.
Adsorption measurements were made at high pressure using an automated volumetric adsorption unit.
Approximately 2-2.5 g of sample was loaded into a stainle6s steel sample cylinder protected with a 20-micron filter to prevent loss of sample. The samples were heated under vacuum at 1~C/min or less to 400~C and held at 400~C until the pressure dropped below lx10-5 ~orr (vacuum activation).
After activation, N2 isotherms were obtained to 12000 torr at 23~~. The isotherms were fit to an expression that imparts empirical heterogeneity to the Langmuir equation:

N~ = P/(a+(bP)+c/(~+(dP)+(dP)2/2+(dP)3/6)) where N~ is the amount adsorbed, P is the pressure and a,b,c and d are coefficients. The fits were used to generate N2 capacities and isothermal N2 working capacities reported in Table IV-A.

- 20 - 2~

TABLE IV-A
Nitrogen Capacity for Mixed Cation (Li,Ca,Na~LSX containing about 85% Li, aft~r Activation to 400~C, ~lx10-5 torr ~
sample Li/Al Ca/Al Na/Al N~(obs)1, N~(delta)2, number eq ratio eq ratio eq rat:io mmol/g mmol/g 1 0.83 0.16 0.00 1.47 1.1~
2 0.81 0.10 0.05 1.28 1.00 3 0.82 0.05 0.10 1.07 0.87 4 0.80 0.00 0.14 0.91 0.77 1Nm(obs~ = nitrogen capacity at 1 atm. and 23~C.~5 2Nm(delta) = isothermal working capacity from 0.2 to 1.2 atm at 23~C.

The e~fect of lithlum in combination with calcium in the presence of other ca~ions, such as sodium, was evaluated by comparing the nitrogen capacities of two series of LSX-zeolite samples containing increasin~ levels of calcium. The flrst set, control samples from Example 2, contained only calcium and sodium, with increasing calcium exchan~e levels of 0, 10, 20, and 30~ calcium. The second set of samples, from Table IV-A, contained approximately 85% lithium, with calclum exchange levels of 0, 5, 10, and 15%, and the balance was sodium. The increasin~
levels of calcium were prepared by displacin~ sodium rather than lithium so that all the samples in the second se~ contained the same amount of li~hlum. The nitro0en capacities at 1 atm, 23~C
of these materials are listed in Table IV-B. Cclumn two contains the nitrogen capacities for the samples containin~ 85% lithlum, and column three contains the nltro~en capacities for the control samples containing no lithium~ Column four contains the differences in nitro~en capacity between each 85% lithiumJ
calcium, sodium LSX-zeolite sample and the nitro~en capacity of 85% lithium, sodium LSX zeolite, and column five contains the differences in nitro~en capacity between each calcium, sodium LSX-zeolite contro~. sample and the nitro~en capacity of pure .. .

:
.. , 2~s~lsr~

sodium LSX-zeolite. It can be observed that calcium displacin~
sodlum in sodium LSX-zeolite has no effect on nitro~en capaclty up to an exchan~e level of at least 30% calclum. The observed increases in nitro~en capacity for the three samples reported in column five, Table IV-B average 0.05 mmol/~ ~reater than sodium LSX-zeolite. Thus, these control samples have the sa~e capacity as sodium LSX-zeollte within experimental error. Unexp~ctedly, calcium displacin~ sodium in 8~% lithium, sodium LSX-zeolit~
si~nificantly lncreases the n~tro~en c,~paclty, by up to 0.56 mmol/~ for 15~ calcium, the maximum calcium that can displace sodium from 85% lithium, sodium LSX-z~olite. Even small amounts of calcium have an effect: 5% calcium increases the nitrog~n capacity by 0.16 mmol/g.

TABLE IV-B
Effect of Ca displacing Na on N2 Capacity at 1 atm, 23~C

Nm, mmol/g Nm difference,1 mmol/g __ __ _ Ca% (85Li,Ca,Na) (Ca,Na)LSX (85Li,Ca,Na) (Ca,Na)LSX
LSXControl LSX Control _ 0 0.910.43 o.ao o.oo 1.07 - 0.16 1.280.53 0.37 0.10 1.47 - 0.56 - 0.49 - 0.06 0.46 - 0.03 1N~ difference = difference in capacity compared to sample with no Ca; i.e., N~ for (85Li, Ca, Na)LSX minus Nm for (85Li, 15Na)LSX; N~ for (Ca, Na)LSX minus N~ for NaLSX.
Additional samples of lithium, calcium, sodium LSX-zeolite were prepared to determine the m; ni calcium and , ~..

22 . 2~8~

lithium exchange levels that are required to observe improved nitrogen capacity of mixed cation lithium, calcium LSX-zeolite over both calcium, sodium LSX-zeolite at the same calcium level and lithium, sodium LSX-zeolite at the same lithium level. The samples were prepared by a~; ng 0.05 M CaC12 and 0.1 M NaCl to lithium LSX-zeolite (#8, #9), or o.05 M CaC12 and 0.1 M LiCl to sodium LSX-zeolite (#5, #6, #7), stirring at room temperature for about 2 h, and then filtering with no wash. Elemental analyses and nitrogen capacities are shown in Table V. The nitrogen capacities at 1 atm. for lithium, calcium LSX-zeolite are compared to nitrogen capacities of :Lithium, sodium LSX-zeolite from Example 1 and calcium, sodium LSX-zeolite from Example 2 at the same lithium and calcium levels, given as reference in the last column of Table V.

.

, 2(~s~sn I'ABLE V
Nitrogen capacity for Mixed cation (Li,Ca,Na)LSX, a~ter Activation to 400~C, ~lxlO 5 torr _ _ _____ __ sample Li/Al Ca/Al Na/AL N~tobs)1, N~
number eq ratio eq ratio eq ratio mmol/g ~re~)2, mmol/g ___ _____ __,.
5 0.07 0.1~ 0.82 0.42 0.47 6 0.12 0.21 0.65 0.42 0.47 7 0.13 0.42 0.45 0.53 0.53 8 0.37 0.33 0.25 0.60 0.~7 9 0.23 0.34 0.40 0.48 0.47 1Nm~obs) = ni~rogen capacity at l atm. and 23~C.
2Nm~ref) = N2 capacity of prior art reference material, either (Li, Na) LSX at the same Li exchange level, or (Ca, Na) LSX at the sAme Ca exchange level, whichever has the greater N2 capacity. ).47 mmol/g is the average N2 capacity for samples of (Ca, Na) LSX containing 30% Ca or le5s.

The nitrogen capacity for mixed lithium, calcium LSX-zeolite is greater than the nitrogen capacities both for lithium, aodium LSX-zaolite and calcium, sodium LSX-zeolite only if the sum of the calcium and lithium ion exchange is greater than about 60%, such as observed for samples 1 to 4 and 8 in Tables IV-A and V.

EXAMPLE~
Calcium, Lithium LSX-Zeolite Binarv N2/02 Selectivities An additional performance factor for air separation adsorbents is ~2 product recovery. Unrecovered ~2 either coadsorbs with N2 or remains in the voids of the adsorben~ bed.

; : ~

~ , .

- 2~

The binary N2/02 selectivity at ~eed pressure is an indicator of the recovery losses from coadsorbed ~2' This example compares the N2/02 selectivi-ties of lithium, calcium LSX-zeolite to the controls, lithlum, sodlum LSX-zeolite of Example 1 and calcium, sodium LSX-zeolite of Example 2.
N2 and ~2 isotherms were obtained to 12000 torr at 23D
and 45~C for the samples described in Example 3 and several control samples from the Examples 1 and 2 using the automated hi~h pressure volumetrlc adsorptlon unlt as described in Example 4. The isotherm data was fit to the Dual Site Lan0muir (DSL) model:
Nm = m1bP/(l~bP) + m2dP/(1+dp) where Nm is the amount adsorbed, P is the pressure , m1 and m2 are the monolayer capacities for sites 1 and 2, respectively, and b and d are the affinity parameters for sites 1 and 2 respectively.
The fits were used.to generate N2 capaci.ties and isothermal N2 working capacities from 0.2 to 1.2 atm at 23~C set forth in Table VI. The trends in isothermal N2 workln~ capacitles from 0.2 to 1.2 atm are the same as those described above in Example 3 for 0.2 to 1.0 atm. Binary N2/02 selectivitles were calculated usin~
ideal adsorbed solutlon theory (IAST) for alr feed at 1.45 atmospheres, 30~C, where N2/02 selectivity is defined as:

~ (N2/~2) = -No2/ Yo2 Where NN2 = N2 coadsorbed at N2 partial pressure in the feed No2 = ~2 coadsorbed at ~2 partial pressure in the feed YN2 = mole fraction of N2 in the feed Yo2 = mole fraction ~~ ~2 in the feed The binary N2/02 selectivities are also included in Table VI.

- 25 ~ 2~

TABLE VI
Nitrogen Capacity and N2/02 Selectlvity for Mixed Cation (Ll,Ca)LSX, (Li~Na)LsxJ and (Ca,Na)LSX after Actlvation to 400~C, <1 x1 o-5 torr ~ ................... -- -- -- _ -- -- _ _ _ . _ -- . -- -- .. _ _ _ . _ sample Ll/Al Ca/Al Na/Al Nm~obs), N~(delta), ~(N2/o2)3 number eq eq eq mmol/~1 mmol/~2 ratio ratio ratio ~ . . .. .

10 1 1.03 n/a n/a 1.46 1.14 10.0 2 0.94 0.10 n/a 1.49 1.13 9.8 3 0.83 0.20 n/a 1.55 1.14 9.8 4 0.73 0.30 n/a 1.57 1.10 g.g 0.59 0.41 n/a 1.51 1.03 10.0 15 6 0.48 0.51 n/a 1.53 1.01 9.9 7 0.28 0.72 n/a 1.59 0.95 10.4 8 n/a 0.97 n/a 1.53 0.82 10.8 9 n/a n/a 1.00 0.47 0.46 3.6 0.70 n/a 0.27 0.49 0.46 4.0 2011 0.83 n/a 0.20 0.75 0.65 5.7 12 nla 0.66 0.30 1.09 0.82 6.9 1Nm(obs) = nitro~en capacity at 1 atm, 23~G.
2Nm(delta) = isothermal workin~ capacity from 0 to 0.3 atm at 23~C.
3a (N2/02) = N2/02 selectivity for air at 1.45 atm, 30~C, calculated from IAST.

FIG 7 compares the blnary N2/02 work~n~ selectivities for lithium, calcium LSX-zeolite to those for lithlum, sodium LSX~
zeolite. The selectivity for lithium, calcium LSX-zeolite is higher ~han that for lithium, sodium LSX-zeolite at ~he same lithium exchange levels.
FIG 8 compares the binary N2/02 selectlvities for lithium, calcium LSX-zeolite to those for calcium, sodium LSX-zeolite. The selectivity for lithium, calcium LSX-zeolite is higher than that for calcium, sodium LSX-zeolite at the same calcium exchange level.

. - . . .... .
:~, . ~ .
' ~ ~

!
.
.

- 26 - ~ '7'~

Thus the binary lithium, calcium ion exchan0ed LSX-zeolites of the present invention exhibit higher N2/02 selectivlty than that observed for the prior art l:Lthium, sodium LSX-zeolite at the same lithlum level and hlgher than that observed for the prior art calcium, sodium LSX-zeolite at the sa~e calcium level.

Lithium, Calcium LSX-Zeolite 0~ VSA Equilibrium Process 5imulat:Lon 'This example predicts ~2 VSA process performance of lithium, calcium LSX-zeolite adsorben~s based on the equilibrlum properties of the adsorbents.
N2 and ~2 isotherms were obtained to 12000 torr at 23~
and 45~ C for the samples in 1'able VIX, derived from Sample #6 of Table VI and Sample #1, #3 and Sample #4 of Table IV-A, the sample of Example 7~ and a commerclal 5A-zeolite, respectively, using the automated high pressure volumetric adsorption unit described in Example 4.
The ~2 isotherms were fit to the Langmuir model with heat effec~s; ie. b = bo exp (Q/RT) where bo is the affinity parameter at infinite temperature and Q is the temperature dependance of the affinity parameter, and the N2 isotherms were fit to the Sircar-Jovanovic model:
Nm = m(1~ PK/mb)'b) K = KOexp(Qk/~T) b = bOexp(Qb/RT) where Nm is the amount adsorbed, P is the pressure, T is the absolute temperature, R is the gas constant, m is the maximum loading, K0 is the Henry's constant at infinite temperature, Qk is the temperature dependence of the Henry's constant, bo is the heterogeneous parameter at infinite temperature, and Qb is the temperature dependence of the heterogeneous parameter. Binary equilibria were calculated using ideal adsorbed solution theory (IAST) described by A. L. Meyers and J. M. Prausnitz in the American Institute of Chemical Engineers Journal, vol. 11, p.121.

.

- 27 - 2~

A computer process model was used to simulate a standard ~2 VSA process cycle at chosen pressures and end of feed temperature. The model is based on ~lobal equilibrium; i.e., it assumes no spatial concentration ~radients and complete bed utilization, and is referred to as GEM. Multicomponent equilibria are estimated by IAST, and heat effects are included.
Input for the program lnclude isotherm<s for N2 and ~2 at two temperatures, and adsorbent physical properties (bul~ denslty, void volume, and heat capacity).
The three major performance fac~ors obtained from the simulations were Bed Size Factor (BSF), Recovery of ~2~ and Actual Cubic Feet evacuated/lbmol Evacuation ~as (ACF/Evac).
Lower BSF, higher Recovery, and lower ACF/Evac indicate improved performance.
Table VII contains the results of the process simulations for an ~2 VSA process cycle with a feed pressure of 1000 Torr and an evacuation pressure of 300 Torr for several calcium, lithium LSX-zeolites, a sodium3 lithium LSX-zeolite control, and a typical commercial 5A zeollte used for air separation. The ~ecovery, BSF, and ACF/Evac for each adsorbent are normalized to a value of 1.0 for the commercial 5A zeolite.
At 75~F end of feed temperature, the lithiuml calcium LSX-zeolites perform significantly better than the commercial 5A
zeolite. The lithium, calcium LSX-zeolites have signlficantly higher recovery and lower BSF than the commercial 5A zeolite, and only moderately hi~her ACFlEvac. At 75~F, the lithium, calcium and lithium, calcium, sodium LSX-zeolites containing 85% lithium perform better than the lithium, sodium LSX-zeolite control containing 85% lithium. They have equal to better recovery, lower BSF, and only minimally higher ACF/Evac. At 105~F, the lithium, calcium and lithium, calcium, sodium LSX-zeolites containing 85% lithium perform significantly ~etter than the lithium, sodium LSX-zeolite control containin~ 85% lithlum. They have higher recovery, significantly lower BSF, and only minimally higher ACF/Evac.

:
.:

- 28 - 2 a ~ 5 TABLE VII
~2 VSA GEM Process Simulations Sample Sample End of Relative Relative Relative 5 Number Identity F~ed Recovery BSF ACF/Evac Temp,F
_ _ _ _ _ .
1 Commercial 5A 75 1.00 1.00 1.00 2 50%(Ca,Li)LSX 75 1.15 0~72 1.05 10 3 33%(Ca,Li)LSX ~clay 75 1.19 0.70 1.05 4 15%(Ca,Li)LSX 75 1.~50.59 1.08 (85Li,lONa,05Ca)LSX 75 1.24 0.64 1.06 6 15~(Na,Li)LSX 75 1.2~0.65 1.05 4 15%(Ca,Li)LSX 105 1.300.61 1.15 15 5 (85Li,lONa,05Ca)LSX 105 1.27 0.71 1.12 6 15~(Na,Li)LSX 105 1.240.78 1.10 Lithium, Calcium LSX-zeolite ~2 VSA E~uillbrium Process Simulation Effect of Temperature A lithium, 33% calcium LSX-zeolite was prepared.from sodium, potassium LSX-zeolite powder derived from clay by addlng lithium LSX-zeolite to a stoichiometric amount of 0.~ molar 25 calcium chloride, heating at 109~C for 16 hJ fllterin~, and washing with water. Elemental analysis indicated a lithium/aluminum equivalent ratio of 0.60 and a calcium/aluminum equivalent ratio of 0.34.
The effect of temperature on process performance was evaluated for the adsorbent using the methods described in Example 6. ~2 VSA performance was simulated at end of feed ~emperatures from ~5 to 135~f. Table VIII contains the recovery, BSF, and ACF/Evac normalized to the results at 55~F. Lith1um, calcium LSX-zeolite demonstrates the unexpected beneficial effects of higher recovery at temperatures up to 135~F and lower BSF at tempera~ures up to greater than 95~F in ~2 VSA. This example demonstrates an added benefit of the li~hium, calcium ~SX-zeolites of the present inven~ion in that they do-not display . , . ~ .

.
.

.,.
:
. ~

.

- 29 - 2~

the deterioration in performance with lncreasin~ tPmper~ture observed for the prlor art lithium, sodium LSX-zeolites in 02 VSA
processes. For the family of lithium, calclum L~X-zeoli~es, those with higher levels of calclum would appear to benef1t more from hi~her temperatures.

TABLE VIII
~2 VSA GEM Proces~ Si.mulation~:
Effect of Temperature on 33% Calci.um, Lithium LSX-Zeolite _ _ _ .Temp, Relative Relative Relative ~F Recovery BSF ACF/Evac _______ _ __ __ 55 1.00 1.00 1.00 15 75 1.05 0.95 1.07 95 1.08 0.94 1.13 115 1.10 0.97 1.16 135 l.lO 1.07 1.18 Lithium, Calcium LSX-Zeolite 0~ VSA E~ullibrium Process Simulation Effect of Dilutina the Zeolltic Phase Although the results presented above were obtained usin~
zeolite powders, it is understood ~hat in a typical PSA process one would use some sort of formed material, such as pellets or beads. Since the forming process often requires the use of an inert binder to provide adequate physical stren~th and attrition resistance, it is important to determine the effect of binder on performance.
The performance of 85% lithium, calcium LSX-zeolite (Sample #1 of Table IV-A) was simulated usin~ the GEM model described above in Example 6. The material was evaluated with no binder, a typical formulation of 80% zeolite/20% binder, and 49%
zeolite/51% binder. The presence of binder was accounted for by multiplying the M's (monolayer coverages) in the isotherm fits by the percent zeolite. The formulations were evaluated at an end of feed temperature o~ 75~F, and Table IX contains ~he resultin0 , , . ~ . ...

- 30 ~

Recovery, BSF, and ACF/evac. As expected, BSF increases with increasing binder, but much less than expected i~ ~here were no benefits for dilution. The expected increases in BSF are 25% for 20% binder and 100% for 50% binder, whereas the observed increases in BSF are 10% For 20% binder and 45% for 51% binder.
Even more unexpected ls the increase in recovery and the decrease in ACF/lb mol evac wi~h increasin~ binder.
TABLE IX
GEM Predictions for 85% Lithium, Calcium LSX-zeolite Containin~ Different Amounts of Binder.
.
% Binder Temp @ Relative Relative Relatlve Relative end of Delta T Recovery BSF ACF/Evac Evac (~F) Binderless 55.1 1.000 ~.000 1.000 1.000 20% Binder 57.7 0.873 1.019 1.104 0.977 51% Binder 62.6 0.623 1.024 1.452 0.~28 These results were to~ally unexpected based on the prior art. Those skilled in the art have generally believed that the addition of binder reduces the adsorptive properties of zeolitic materials. In the past, the trend has been to try to reduce the levels of binder from the typical 20% to as low as possible, often as low as 5%, or even to eliminate the binder entirely.
This example demonstrates that it is preferred to use these materials diluted (with binder) in standard ~2 VSA cycles.

:

2~8~

Lithium, Strontium LSX-Zeolite N2Capacities Strontium LSX-zeolite was prepared by ion exchange of sodium, po~assium LSX-zeolite using four static exchan~es at 100~C
with a 3.8-fold equivalent excess of 1.0 M Sr(NO3)2 adjusted to a pH of about 7 using Sr(OH)2. Two samples of lithium, strontium LSX- zeolite were prepared by addin~ lithlum LSX-zeolite powder to various amounts of 0.05 M Sr(NO3)2 ad~usted to a pH of about 7 using 0.1 M LiOH and stirring at room temperature for about 4 h.
The mixçd cation samples were filtered, but not washed. Table X
contains the results o~ elemental analyses for strontium and lithium and N2 capacities and binary N2/02 selectivities obtained as described in Example 5.
Fi~. 4 is a graph of nitrogen capacity at 1 atmosphere and 23~C for the extent of lithium exchange in lithium, strontium LSX-zeolite. It shows that capacity uniformly and unexp~ctedly is in excess of what might be expected from straight line extrapolation of 100% strontium to 100% lithlum ion content.
As observed for the mixed lithium, calclum LSX-zeolites, the N2 capacities, isothermal nitrogen working capacities and binary N2/02 selectivities for mixed lithium, strontium LSX-zeolites are si~nificantly higher than those for lithium, sodium LSX-zeolites at the same li~hium exchan0e level. Thus dlvalent alkaline earth metal cations other than Ca2+ also show unexpectedly high capacity in admixture with lithium.

', - .

, , ~ .
:

- 32 - ~ ~8~1~0 TABLE X
Nitro~en capacity and N2/02 Seleotivity for Mixed Cation (Li,Sr)LSX after Activation to 400~C, ~lX10 5 torr ~ .
sample Li/Al Sr/Al N~(obs), N~(delta), ~(N2/o2)3 nu~ber eq ratio eq ratio mmol/~1 mnol./~2 ..
1 1.03 n/a 1.46 1.14 10.0 10 2 0.77 0.20 1.41 1.08 8.91 3 0.66 0.30 1.34 1.03 8.32 4 n/a 1.07 0.98 0.77 5.85 1Nm~bs) = nitr~en capacity at 1 atnl. 23~C.
2Nm(delta) = isother~al workin~ capacity from 0.2 to 1.2 atm 3a ~N2/02 = binary N2/02 seleotivity for air at 1.45 at~, 30~C.

Llthlum, Potassium LSX-Zeolite Control Several samples of lithlum, potassium LSX-zeolite were prepared by addin~ lithium LSX-zeolite powder to varyln0 amounts of 0.1 M KCl and stirring at room temperature for about 4 h. The samples were filtered but not washed. Table XI contains the results of elemental analyses for lithium and potassium and nitrogen capacities obtained as described in Example 4.

:

, ., .

- 33 ~ ?J~ 8 TABLE XI
Nitrogen Capaci~y for Mixed Cation (Li,K)IsX containing about 85% Li, after Activation ~o 400~C, <lxlO-5 torr - - .
sample Li/Al K/Al N3(calc), N~(obs), number eq ratio eq ratio mmol/gl mmol/g2 - l 1.03 0.00 ' 1.35 1.35 2 0.87 0.10 1.24 0.96 3 0.79 0.19 1.15 0.6~
4 n/a 0.98 0.26 O.Z6 1N~(calc) = 1.354*I.i/(Li~K) + 0.263*X/(Li+X), calculated N2 capacity at 1 atm., 23~C, based on the capacities of the two end members.
2Nm~obs) = nitrogen capacity at 1 atm., 23~C.

2~
The N2 capacity of llthlum, potassium LSX-zeolite decreases si~nificantly with addltion of po~assium to 10 and 20%
levels, similar to the behavior of lithium, sodium LSX-zeolite.
In addition, the observed capacity (Nm(obs)~ is significantly less than what mi~ht be expected (Nm(calc)) from a straight line extrapolation of 100% potassium to 100% lithium content. This control further supports the unique result obtained for lithium in admixture with calcium as compared to lithium in admixture with monoYalent alkali metal cations.

Li~hium. Calc~um, Potassium LSX-Zeolite N~ Capacities Three samples of lithium, calclum, potassium LSX-zeollte were prepared by addin~ lithium LSX-zeolite powder ~o various amounts of 0.1 N solution con~aining Yarious amounts of CaCl2 and KCl and stirrirg at room temperature for about 4 h. The samples were filtered but not washed. Table XII contains the results of elemental analyses of lithium, calcium and potassium and N2 capacity at 1 atm obtained as described in Example 4.

, , .... ~ :

; . . ; .. . : , . .
, :
..
. .
:' TABLE XII
Nitrogen Capacity for Mixed Cation (Li,Ca,K)LSX Containing ahout 85% Lithium after Activation to 400~C, <lxlO 5 torr sample Li/Al Ca/Al K/Al N~(obs),1 number eq ratio eq ratio eq ratio mmol/g . _ _ _ l 0.83 0.16 0.00 1.47 2 0.80 O.lO 0.05 1.22 3 0.82 0.05 O.lO 0.97 1Nm(obs) = Nitrogen capacity at l atm., 23~C.

Calcium displacing potassium in 85~ lithlum, potassium LSX-zeolite significantly increases the N2 capacity. For comparison, 85~i/15K uptake is 0.78 mmol/g, determined by interpolatin~ between values ~or 80Li/20K and 90Li/10K, the control samples in Example 10. Even small amounts of calcium have an effect: 5% calcium increases the N2 capacity by 0.19 mmol/g from 0.78 mmol/g to 0.97 mmol/~. This example demonstrates the beneficial effect of lithium in combination with calcium for other ternary ion exchanged forms that contain potassium instead of sodium in addition to lithium and calcium.
EXAMPLE1~
Lithium, Calcium and Lithium. Strontium X-~eolite N2 CaDaoities Lithium X-zeolite was prepared from Linde 13X (sodium X-zeoliteJ using five static exchanges at 100~C with a 5.6-fold equivalent excess of 1.1 M LiCl. Two samples of lithium, calcium X-zeolite and one sample of lithium, strontium X-zeollte were prepared by adding lithium X-zeolite powder to either 0.05 M
CaCl2 or 0.05 M Sr(N03)2, respectively, and stirrin~ at room temperature for about 4 h. The mixed cation samples were flltered, but not washed. Table XIII contains the results of elemental analyses for lithium, calcium and strontium and N2 capacity at 700 torr obtained as described in Example ~.

~ - ~ : , :

: . ~: . ..
.

;:
.

..

~ 35 ~ 2 ~ 8 0 TABLE XIII
Nitrogen Capacity for Mixed Cation (Li,Ca)X and (Li,Sr)X after Activat:ion to 400~C, <lxlO 5 torr ___ _ 5 sample Li/Al Ca/Al Sr/Al Na/Al N~(obs~, number eq ratio eq ratio eq rat:io e~ ratio mmol/g1 .. _ l 0.98n/a n/a 0-04 l.00 2 0.700.22 n/a 0.04 0.94 3 0.520.37 n/a 0.03 0.96 4 0.68n/a 0.34 0.03 0.88 1N~(obs) = Nitrogen capacity at 0.9 atm., 23~C.

15FIG 9 compares the N2 capacity at 700 torr, 23~C, for the lithium, calcium and lithium, s~ron~ium binary ion exchan~ed forms of X-zeolite powder to data presented for the llthium, sodium binary ion exchanyed forms of "binderless" X-zeolite in US
~,~59,217. The N~ capacities for lithium, calcium X-zeolite and lithium, strontium X-zeolite are higher than the N2 capacities for lithium, sodium X-zeolite at the same lithium exchan~e level.
This example demonstrates that the unexpected result observed for X-zeolite containin~ a Sl/Al=1 (LSX-zeolite) is also observed for X-zeolite at higher Si/Al ratios such as 1.2. It also supports Example 9 in that i~ demonstrates that divalent alkaline earth metal cations other than Ca2+ also show unexpectedly high capacity in admixture with lithium.

Lithium Calcium LSX-Zeolite Extrudate Flow Activation A sample of lithium, calcium LSX-zeolite extrudate was prepared by slx statlc ion exchanges of calcium, sodium I~SX-zeolite extrudate with 2.~ M LiCl at 100~C. Two flow activation experimental runs were performed in the followin0 manner. A 30 cc portion of the lithium, calcium exchanged extrudate was placed in a 1-in diameter stainless steel sample cylinder, which was placed in a tube furnace. In order to activate the sample, gas flow was initiated through the sample, and the furnace -' ~

.
, .

, - 36 . 2~

temperature was controlled with a programmable temperature con~roller (flow activation). Two 30cc sample portlons were flow activated as follows:
Run #1: N2 at 1.3 L/min, heated at 2~C/min to ~00~C and held at 400~C for 4 h. Final dew point of exit gas = -20~C.
Run #2: Step 1: Breathin~ air (contains C~2) flowin~ at 2.6 L/min saturated with H20 at room temperature, heated at 10~C/min to 120~C and held for 2 h 40 min.
Step 2: N2 at 1.3 L/min, heated at 10~C/min to ~00~C and held 400~C for 4 h 30 min. Flnal dew point of the exit ~as = -45~C.
At the end of the sample activation, valves at the ends of the sample cylinder were closed, the sample was allowed to cool, and then evacuated. N2 isotherms to 100 psic~ were obtained on a volumetric isotherm unit at 30~C controlled with an air temperature bath. A third portion of the extrudate was vacuum activated as described in Example 4. rhe nitro~en capacities at 1 atm and 30~C for the flow activated samples were compared below to that obtained on the volumetric isotherm unit for the vacuum activated sample.
Sample N2 Capacity at 1 atm, 30~C
~ . . ~ . ..
vacuum activated 0.93 mmol/~
flow activated (run #1) 0.90 mmol/~
flow activated (run #2) 0.91 mmol/~
This example demonstrates that mixed cation lithium, calcium LSX-zeoli~e adsorbents can be activated ei~her by vacuum or in the absence of vacuum, provided that the gas composition, ~low rate and temperature ramp are controlled to limit the presence ~f H20 ancl C02 in the atmosphere.

, , . .
;

8 ~

Lithium, Ca].cium LSX-Zeolite Effect Of Order Of Excha~ae Of Cations On N~ Capacity A sample of lithium, calcium lSX-zeollte was prepared by lithium ion exchange of calcium LSX-zeolite powder usin~ six static exchanges at 100~C with a 6.3-fold equivalent excess of 2.2 M LiCl. Elemental analysis of the sample gave a Li/Al equivalent ratio of 0.70 and a Ca/Al equivalent ratio of 0.25.
The following N2 capacities at 23~C were obtained as described in Example 4:
Nm(obs): 1.33 mmol/g Nm(delta): 0.98 mmol/g Thus lithium, calcium LSX-zeollte prepared by lithium exchange of calciu~ LSX-zeolite shows the same improvement in adsorptive properties compared to lithium, sodium LSX-zeolite and calcium, sodium LSX-zeolite, as demonstrated by lithium, calclum LSX-zeolite prepared by calclum exchange of lithium LSX-zeolite.
The method of ion exchan~e ls not limited to the procedures described above. The same compositlons prepared by other ion exchange routes should perform essentially the same as the materials reported herein.
The lithium, calcium X-zeolite adsorbents of the present invention exhibit some unexpected and remarkable performance characteristics when used to selectively adsorb nitro~en ~rom ~as mixtures containing nitrogen in contrast to other adsorbents containin0 lithium or calcium used for suoh nitro~en adsorption processes. In particular, the N2 capacity of mixed cation lithium, calcium LSX-zeolite exceeds what might be expected from a strai~ht line extrapolation of the capacities of the end members. This unexpected result is in marked contrast to the relevant prior art materials, calcium, sodium LSX-zeolite or lithium, sodium LSX-zeolite. In addition, both the nitrogen working capacity and the nitrogen/oxy0en selectivity of the at least binary ion exchan~ed X-zeolites of ~he present invention are higher than those observed for the prior art lithium, sodium X-zeoli~es at the same lithium level and hi~her than those observed for the prior art calcium, sodium X-zeolites at the same .

~~8~~

calcium level. Even small amounts of calcium have a si~nificant effect, as observed by significant increases in N2 capacity as a result of displacing sodium with calcium in lithlum, sodium LSX-zeolite, as compared to no chan0e in N2 capacity as a result of displacin~ sodium with calcium in sodium LSX-zeolite. In addition to the improved adsorptive properties of the adsorbents of the present invention, these materlals exhibit some unexpected performance in ~2 VSA process simulations. Specifically, VSA
performance improves with increasin~ temperature above ambient, whereas the prior art lithium, sodium X-zeolites deteriorate in performance with increasing temperature. VSA performance oF
lithium, calcium X-zeolites also surprisingly lmproves with dilution of the zeolitic phase to levels above those typically used for binding zeolites for granulation purposes.
The present invention has been set forth with reference to several preferred embodiments. However, the full scope of the invention should be ascertained from the claims which follow.

E: \GLC\4748APLD . 002 , '

Claims (19)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE AS CLAIMED ARE DEFINED AS FOLLOWS:

1. A process for selectively adsorbing nitrogen from a gas mixture containing nitrogen and at least one less strongly adsorbed component which comprises contacting the gas mixture with an adsorbent which is selective for the adsorption of nitrogen, comprising a crystalline X-zeolite having a zeolitic Si/Al ratio ~ 1.5 and an at least binary ion exchange of the exchangeable ion content with between 5% and 95% lithium and with between 5% and 95% of a second ion selected from the group consisting of calcium, strontium and mixtures thereof, wherein the sum of the lithium and second ion ion exchange is at least 60% of the exchangeable ion content.

2. The process of Claim 1 wherein the zeolite is ion exchanged with lithium to approximately 50% to 95%.

3. The process of Claim 1 wherein the zeolite is ion exchanged with the second ion to approximately 5% to 50%.

4. The process of Claim 1 wherein the zeolite is ion exchanged with approximately 15% of the second ion and 85% lithium.

5. The process of Claim 1 wherein the second ion is calcium.

6. The process of Claim 1 wherein the second ion is strontium.

7. The process of Claim 1 wherein the zeolite is ion exchanged with approximately 15% calcium and 85% lithium.

8. The process of Claim 1 wherein the gas mixture contains nitrogen and oxygen.

9. The process of Claim 1 wherein the gas mixture is air.

10. The process of Claim 1 wherein the Si/Al ratio is approximately 1.

11. The process of Claim 1 wherein an oxygen and nitrogen containing gas mixture contacts a zone of said adsorbent, the nitrogen is selectively adsorbed and the oxygen passes through the zone and is recovered as an oxygen enriched product.

12. The process of Claim 11 wherein the oxygen product has a purity of at least approximately 90% oxygen.

13. The process of Claim 11 wherein the adsorption is conducted at an average bed temperature in the range of approximately 55 to 135°F.

14. The process of Claim 11 wherein the zone is operated through a series of steps comprising: adsorption during which the gas mixture contacts the adsorbent, nitrogen is selectively adsorbed and oxygen passes through the zone as a product; depressurization during which the gas mixture contact is discontinued and the zone is reduced in pressure to desorb the nitrogen; and repressurization with oxygen product to the adsorption pressure.

15. The process of Claim 14 wherein the adsorption pressure is in the range of approximately 35 to 65 psia.

16. The process of Claim 14 wherein the desorption is conducted down to a pressure in the range of approximately 14.7 to 16.7 psia.

17. The process of Claim 11 wherein the zone is operated through a series of steps comprising: adsorption during which the gas mixture contacts the adsorbent, nitrogen is selectively adsorbed and oxygen passes through the zone as a product; depressurization during which the gas mixture contact is discontinued and the zone is reduced in pressure to desorb the nitrogen; evacuation to further desorb the nitrogen to below ambient pressure; and repressurization with oxygen product to the adsorption pressure.

18. The process of Claim 17 wherein the adsorption pressure is in the range of approximately 900 to 1600 torr.

19. The process of Claim 17 wherein the evacuation is conducted down to a level in the range of approximately 80 to 400 torr.